1. Field of the Invention
The present invention relates to a method and an apparatus for fabrication of a specimen. More particularly, the present invention relates to a method and an apparatus for extracting a micro-specimen including a specific small area of a semiconductor material such as a semiconductor wafer or a semiconductor device chip from the semiconductor material by separation using an ion beam and for fabricating a specimen used for carrying out an observation, an analysis and/or a measurement of the specific small area.
2. Description of the Prior Art
In recent years, efforts made to shrink geometries of semiconductor devices make progress at a very great pace. In a structure analysis of these semiconductor devices, there has been demanded an observation of a nanoscopic structure which is so small that, at a resolution of an ordinary scanning electron microscope referred to hereafter simply as an SEM, the structure can not be observed any longer. As a result, observation by means of a transmission electron microscope which is abbreviated hereafter to a TEM is indispensable in place of an SEM. Traditionally, however, fabrication of a specimen for an observation using a TEM can not help resorting to manual work which must be done by a well trained person and takes a long time. For this reason, in reality, the method for observation of a specimen using a TEM does not come into wide use as the method for observation by means of an SEM, whereby a specimen can be fabricated with ease and results of observations can be thus be obtained immediately, did.
The conventional method for fabrication of a specimen for an observation by using a TEM is explained as follows. FIG. 2 is diagrams showing the first conventional method for fabrication of a specimen for observation using a TEM. A specimen for observation using a TEM is also referred to hereafter simply as a TEM specimen. To be more specific, FIG. 2/(a) is a diagram showing a semiconductor wafer 2 on which LSIs were fabricated. The semiconductor wafer 2 is referred to hereafter simply as a wafer or a substrate. As shown in FIG. 2/(b), the wafer 2 comprises an upper-layer portion 2A and a lower-portion 2B or a substrate. Assume that a specimen for TEM observation of a specific area on the wafer 2 is fabricated. First of all, a mark not shown in the figure is put on an area 22 subjected to the observation using a TEM. By exercising care so as not to damage the area 22 to be observed, an injury is deliberately inflicted on the wafer 2 by using a tool such as a diamond pen in order to cleave the wafer 2 or the wafer 2 is cut by means of a dicing saw in order to take out a sliver chip 21 shown in FIG. 2/(b). In order to make the center of a TEM specimen being created the area 22 to be observed, the areas 22 of two chips are stuck to each other by using adhesive 23 to produce 2 specimens 24 stuck together as shown in FIG. 2/(c). Then, the two stuck specimens 24 are sliced by means of a diamond cutter to produce slice specimens 25 shown in FIG. 2/(d). The dimensions of each of the slice specimens 25 are about 3 mm×3 mm×0.5 mm. Then, the slice specimen 25 is put on a grinding plate to be ground by using abrasives into a thin specimen, namely, a ground specimen 25′ with a thickness of about 20 microns. Subsequently, the ground specimen 25′ is attached to a single-hole holder 28 mounted on a TEM stage, that is, a stage for holding a TEM specimen as shown in FIG. 2/(e). Then, ion beams 27 are irradiated to the surfaces of the ground specimen 25′ as shown in FIG. 2/(f). Sputtering fabrication (or ion-milling fabrication) is then carried out on the center of the specimen 25′ as shown in FIG. 2/(g). Finally, when a hole has been bored through the center of the specimen 25′, the irradiation of the ion beams 27 is halted as shown in FIG. 2/(h). A thinned area 26 with a thickness not exceeding a value of about 100 nm fabricated as described above has been observed by a TEM. This method is described in references such as a book with a title of “High-Resolution Electron Microscope: Principle and Usage”, authored by Hisao Horiuchi and published by Kyoritsu Syuppan, Page 182, and used as prior-art reference 1.
FIG. 3 is a diagram showing the second conventional method for fabrication of a TEM specimen. This method is a method for fabrication of a specimen using a focused ion beam which is abbreviated hereafter to an FIB. As shown in the figure, first of all, a mark not shown in the figure is created by using a laser beam or an FIB in the vicinity of an area 22 to be observed on the wafer 2 and then the wafer 2 is diced as shown in FIG. 3/(a). A sliver chip 21 shown in FIG. 3/(b) is then taken out from the wafer 2. The sliver chip 21 is further sliced to produce slice specimens 21′ shown in FIG. 3/(c). The dimensions of each of the slice specimens 21′ are about 3 mm×0.1 mm×0.5 mm which is the thickness of the wafer 2. Then, the slice chip 21′ is ground into a thinned specimen 21″. The thinned specimen 21″ is then stuck to a TEM-specimen holder 31 which resembles a thin metallic disc plate and has a out portion 31′ as shown in FIG. 3/(d). Subsequently, the area 22 to be observed on the thinned specimen 21″ is further thinned by means of an FIB 32 so that only a slice 22′ having a thickness of about 100 nm is left as shown in FIG. 3/(e), (f). The slice 22′ is used as a specimen for an observation using a TEM. This method is described in documents such as a collection of theses with a title of “Microscopy, of Semiconducting Materials 1989”, Institute of Physics Series No. 100, Pages 501 to 506, which is used as prior-art reference 2.
FIG. 4 is a diagram showing the third conventional method for fabrication of a TEM specimen. The method is disclosed in Japanese Patent Laid-open No. Hei 5-52721 which is used as prior-art reference 3. As shown in the figure, first of all, a specimen substrate 2 is held in such a posture that an FIB 32 is irradiated to the surface of the specimen substrate 2 perpendicularly. The surface of the specimen substrate 2 is then scanned by the FIB 32 along the circumference of a rectangle to form a rectangular hole 33 with a sufficient thickness on the surface as shown in FIG. 4/(a). Then, the specimen substrate 2 is inclined so that the surface thereof forms a gradient of about 70 degrees with the axis of the FIB 32 and a bottom trench 34 for separation is further created on a side wall of the rectangular hole 33 as shown in FIG. 4/(b). The gradient angle of the specimen substrate 2 is adjusted by using a sample stage which is not shown in the figure. Subsequently, the orientation of the specimen substrate 2 is restored to its original posture so that the FIB 32 is again irradiated to the surface of the specimen substrate 2 perpendicularly and a trench 35 is further created as shown in FIG. 4/(c). Then, by driving a manipulator for holding a probe 36, the tip of the probe 36 is brought into contact with the surface of a portion 40 of the specimen substrate 2 to be separated as shown in FIG. 4/(d). It should be noted that the manipulator itself is not shown in the figure. In this state, the FIB 32 is irradiated to a local area including the tip of the probe 36 while gas 39 for deposition is being supplied from a gas nozzle 37 to create an ion-beam-assisted-deposition film 38 which is abbreviated hereafter to an IBAD film or a deposition film. In this way, the portion 40 of the specimen substrate 2 to be separated and the tip of the probe 36 which have been brought into contact with each other are firmly joined to each other by the deposition film 38 as shown in FIG. 4/(e). Finally, portions left around the portion 40 of the specimen substrate 2 to be separated are separated by the FIB 32 to detach the portion 40 from the specimen substrate 2 as shown in FIG. 4/(f). The detached portion 40 separated from the specimen substrate 2 remains in a state of being firmly joined to the tip of the probe 36 as shown in FIG. 4/(g). An area on the separated portion 40 to be observed is further thinned by using an FIB to a thickness of about 100 nm to produce a specimen for observation using a TEM.
The first and second conventional methods described above can not help resorting to manual work requiring skills of a well trained person fabricating the specimen. The manual work includes grinding, mechanical fabrication and sticking the specimen to the TEM-specimen holder. In addition, with these conventional methods, in order to fabricate a desired specimen, it is necessary to split the wafer or the substrate of the device chips into portions by cleaving or cutting the wafer or the substrate. In order to acquire a specimen of a desired area, portions adjacent to the desired area are inevitably and/or inadvertently cleaved or cut. Assume that it is necessary to observe and/or analyze a portion other than an area which was subjected to an observation and/or an analysis before. Since the substrate of the specimen was once cut in order to fabricate specimens for the prior observation and/or analysis, an injury and/or a damage was inevitably and/or inadvertently inflicted upon the portion subjected to the next observation and/or analysis or a positional relation among portions to be observed and/or analyzed is no longer known. As a result, there is raised a problem that accurate information on observations and/or analyses can not be obtained continuously due to the inflicted injury and/or damage. In addition, while the ion milling and the process to thin a film by using an FIB described above do not directly involve manual work, they have a problem of a long fabrication time which is difficult to solve.
Furthermore, in recent years, there is seen a trend of an increasing wafer diameter to 300 mm. The number of device chips that can be fabricated from such a wafer also increase as well. In addition, the device itself has more added values. As a result, splitting a wafer into portions by cleaving or cutting the wafer in order to observe and/or analyze a particular area leads to a disposal to discard portions other the area to be observed and/or analyzed which is very uneconomical. Moreover, when a small particle or an abnormal shape is detected in a certain area during a scanning operation over the entire wafer by driving a variety of microscopes, a cause of such a small particle or such an abnormal shape has to be clarified by conducting an observation and/or an analysis prior to the splitting a wafer into chips, in particular, before the small particle disappears. Otherwise, a number of defective devices among final products will be resulted in, incurring an even larger loss. If a plurality of specimens can be produced in a short period of time without splitting the wafer into portions, observations and/or analyses can be carried out very economically, giving rise to a great contribution to improvements of a product manufacturing yield.
With the third conventional method, on the other hand, once a specimen is set on the sample stage, it is not necessary for the operator to do manual work directly till separation of micro-specimens and to cut the wafer carelessly. In this method, however, the separated specimen remains in a state of being attached to the tip of a probe so that, when the separated specimen is brought into an observation apparatus and/or an analyzer in such a state to be observed and/or analyzed, the specimen will vibrate, raising a problem that it is impossible to obtain reliable results of observation and/or analysis.
As the conventional TEM-specimen holder, a holder 78 with a single hole 79 shown in FIG. 7/(a), a holder 80 with a notch 108 shown in FIG. 7/(b) and a holder 109 with a mesh shown in FIG. 7/(c) are known. Assume that the single-hole-type holder 78 or the notch-type holder 80 is used in the third conventional method for specimen fabrication described above to hold a micro-specimen 40 with a small size in the range 20 to 30 microns. In this case, it is necessary to adjust the position of the micro-specimen 40 on the inner wall of the notch 108 or the single hole 79 with a high degree of accuracy, making the installation work difficult to carry out. Such a problem is not encountered with the mesh-type holder 109. This is because, by using a mesh-type holder 109 with a gap between mesh nodes adjusted to the size of the micro-specimen 40, the position at which the micro-specimen 40 is to be installed can be selected arbitrarily to a certain degree. With the mesh-type holder 109, however, an electron beam Path 82 propagating toward an area 81 to be observed is shielded by a mesh structure member 109′ as shown in FIG. 7/(d), making an observation using a TEM impossible in some cases.